.

..
m-

SocleW=f

PetroleumEnglnaws

SPE 28008
Use of Transient Testing in Reservoir Management
M.M. Kamal, D.G. Freyder, and M.A. Murray, ARCO E&P Technology
SPE Members

CWWM

1S94

Thispiperwas

society
ofPelroleum
prepared

Engine&,

for presenf.t[m

1..:

i.! !!!. University of Tuba

Ce.l<nnM

pemk.m

Eriglneer!nq

timposi.m

held i. Tulsa, OK, U.SAy

2941.

AUWS4 1994,

was wlwted
for presma! (on by an sPE Prcgrm! COnmmlee following review of intmmatron comaiflnd in an abslraot mbm!md ~y !he author($), mnte.f~
of the paper,
This P&r
., Presented. have 0! been Wviwed Ly the %dely ~ P.4,c,I,w
@#nea<s and am subjti
to COKW$W by tie auth.x(e), Tiia m!erial, as P,esmied, does M e,essarlly ,EW3
any P.M..
of the Society of Pe!role.rn Engrnear& its O.IUCWS,or membws, Papain prasenjed al SPE medinga are wbjea to p.blicai(on review by Ed(mrial Commillm,
of lhe Smlely
should WTtain c.nspicws
a.hmwkd+?me.l
.1 Petroleum E.gl.eers. Permission to copy 19mwct.d
m a. abstract 01 W! mm. than 3C0 wtis,
111.skation$ may not b. w+ad. The
of where and by whom the PaPW !S Presented, Wrlle Lit.raflan, SPE, ?.0. Box ZZW36, Richardson. TX 7508%Z838. USA. TAx,
1SZ245 SPEUT.

abs!re.c!

Abstract
Tfansient testz cart be used throughout the life of the reservoir to
manage and optinize the recevery of hydmarbon fluids. This
paper shows applicatimrs of various well testing methods during
the exp]omtion, appraisal, primary recqye~ ~d SKO!I%.
recovery phases of field projects.
Single-well .Estz (e.g; drills@m.. tests and. buildup tests) are
mostly used ducing @e various.stages of field development and
prirnarj-recovefj+ whereas muftiple-well testz (e.g.: interference
tests) are most often used duiirrg secondaty recovery. A clear
understanding of ihe type of irifomation obtained from weII tasta
is essemial if this infonnerionis to be used properly in managing
the reservoir. Examples are tie dtiferent vaIuei of penneabifities
calculated from single verticzf -well teata, horizontal-well tests
and multiple welf teatz, the vafues of average reservoir prexsures
end the double porosity parameters calculated from t+k in naturally fractured reservom. The staf+of-the-arr of testingand measwernent twls, acquisition iyskms and interpretation me@odS.
affect the type of information obtained from weff tests md aboufd
be considered by the petroleum engineer before using the rezults.
The paper presents informatiori to help the engineer use transient
tests properly end to their fullest extent. Field examples ae nzed
when appropriate.

The nitin difference betweei@tioleti

&~ineering arid most of
the otfm engineering dizeiplincs is rhat the systems mat petrOIeum engineers work witl. am given to them, whereas in otler

519

engineering disciplines the engineers design their systems. For

example, the car engine is designed by a mechzrdcaf engineer and
the structure of a building is designed by a civil engineer. Ofl aid
gas fields are not designed or developed by petroleum engineers.
They are given to them. As a matter of fact, the biggest rlaffenge
psrroleum engineers face, is to characterize tlese systems. fn this
regard, petroleum engineers have the same problem az mhdng
and meteorological enginzera and ground water hydrologists
However, peuoleum engineers have the added complexity of not
being abIe to physically we and dhectfy measure the in-situ prop
erties of our systems. A complexity both mining and meteorologicalenginems do not have. Therefore, petroleum enginmis rely
on induect meaaurementa to obtain the physicaJ properties of
hyckocarkm-beming formations. Examples of theze indwect
meaiurementa are inferring formation penneabifity from production data, calculating porosity from a sonic log or determining the
distance to a fardt from a buildup teat.
fn this paper we discuzs the reserxoir properdea that can be determined tlom various transient testz, how they can be irrtegmted
with other methods m obtain memoir descriptions, and the ma
of this information in rezervoir management. Field examples am
presented ti ilhmtrate the practicality of these techniques.
Tomsient tests, or weff tests, we used during the various stages of
maervoir discovery, development, and production. Drillstem tests
and wireline formation tests are run in explomtion and appraisal
wells drawdown, buildup, interference and pulse tests are run
d@rg primary, secondary and enhanecd recovery stag~, and
step rate, ~mjectivity,falloff, interference and pulse tests are run
during secondary and enhanced recovery stagez. Other specialized tesrz like muldfayer and verdcaf perrneabtiity testz are run

b
2 = -. -..

. . ..-

.>..

USE OF TRAIWENT TESTING IN RESERVOIR MANAGEMENT

throughout the life of the reservoir. A fist of the various reservoir/.

well system profwdea that can be obtained horn each test is
shown in TabIe 1.
Table 1: Reservoir Properties that maybe Obtabred from Various
lYansient Tests
Reservoirbehavior
Permeability
W.
Fracturelens!h
Reservoirpmrsure
Reservoirlimit
Boundaries

fkilt StemT+st

RepeauivfuttipleFnrmationTests

Pressnrcprofile

DrawdownTest

Reservoirbehavior
PenneabUhy
Skin
Fraclurelength
Reservoirlimit
Boundaries

BuildupTesk

Formationpartingpressure
Permeability
Skin

FatioffTest- --

Mobilityin variousbanks
skin
Reservoirpressure
Fracturelength
Locationof front
Boundaries

interferenceand PulseTests

Cmirnmnicationbetweenwetfr
Reservoirtypebehavior
Porosity
bk~dr permeabfity
Verticalpermeability

LayeredResemoirT&ts-

Propertiesof indWidustlayers
Horizontalpeoneabitity
Verticalpermeability
skin
Averageiayer pressure
Outerboundties

recent l-woks on this subject. For our pnrpose here, it would snftice to mention that the petroleum engineer examines the test data
in sevemf grapldcd presentations to determine the reservoir
model and flow regimes. These plots include a log-log plot of the
pressnre difference and the pressure derivative vhrs the testing
rim% which ia cafled the diagnostics plot, and severat specirrfiied
P1OL$each serving to identify a specitic flow regime. A fiiring of
the various plots rmd flow regirne$ together with the information
obtained from each one is shown in Table 2.
fn the following aeclions, we discuss in some detaif the valuea of
the parameters obtained from wett tests rmd how tlrey may be
nsed to help describe and manage the reservoir ffrrougbont its
productive life.

Reservoir Model
A mathematical model that closely matches the reservoir bebavior k useful in predicting tbe field performance. Transient testing
is an excellent source of information about the reservoir descrip
tion and the reservoir models that cmr be used m manage oil and
gas fields. Transient tests are nsefnf becanae the flow regimes
encountered during a well teat depend on the cbrtracteristica of
the reservoir/welt system. For example, ifa single hy&aufic frscorre intersects the wellbore, part of Ure data may exhibit linear
flow.Transient tests are the best tool for determining the effective
fracture length and conductivity. Other examples are the distinct
bebaviors of natttmfly fractnred or layered reservoira.

Reswvoicbehavior
Permeability
Skin
Fractnrelength
Reservoirpressure
Boundaries

Step RateTe3S

SPE 28008

One of the reservoir characteristics identified using weft teata, is

the presence of boundades (faults, permeability pinch outs, etc.).
Thk is prrr-ticutarlybrrportant when testing discovety or sppraissJ
wells. The reservoir model is usnrdfy determined from the shape
of the pressnre data. Maulrews and Russell presented examples of
shapes of pressure data, graphed on semi-log plots, for aeverst
reservoir models.7 Ehfig-Economides farer adapted the Matthews
and Russell examples to preasore difference and derivative loglog plors.g & mentioned in tie intmdnction, identifying the reswvoir model represents the drst step of the snatyais. Severrd
strrdies have used pattern recognition or neuraf nelworka to identify reservoir modefa from welf tests.g-i 1 Attemptr were afso
made m help the engineer identify the medel Uwongbvarious sratisticsf methods.12 Resrdta from compnter-aided anafyses to
identify the reservoir model are enconrsging, but more work is
needed before it can be used as a routine anafysis metbcd.

After reviewing and checking the quafity of raw dsa anafysis ofr
welf rests can be divided into two steps. The reservoir model and
the vsrions flow regimes encountered during the teats tie identified in the first step. In the second step, the vafues of various reservoir and well parameters wc Grtctdated. Numerous pubficatiorrs
described how to anafyze wetf tests and calculate reservoir properties. For example, the first six references are some of the most

When more than one model matches the test data, the engineer
should nse otfrm reservoir description methods to &tennine
which is the moat probable mcdel. For example, if the data
nmtcbes a naturally fiactured (double porosity) or a laywed reservoir model, the well logs or coIea may help decision the appro.
priate model. Arrotherexanrple is when a finear boundary and a
520

SPE 28008

.Jd.M. KAMAL, D. G. FREYDER AND M. A. MURRAY

3.

Table 2: PIotz and Flow Reghnezof lYansientTeztz

Flow
Regime

Cartesian

Wellbore
Storage

. SLLine
.Slopeac
. Intercept *
At=orr

.&

4fit

AP
corr
Linear Flow

. St. Line
. slop =
m,f * L,

Log-log

Semi-1og

. unit
Slope on
&&P
. Ap &p mincide

\POsitives
Negatives
f

. slope = 1/20 ,
&On Apifs= t
. Slope <1/2 on Ap
ifs * O

. Intercept *
Fracturedamage

. paat 1/2 level of&

Bitinew HOW

. St. Line,
. Slope = mbf
/

. slope = 1/4

. p at 1/4level of Ap

F=
FirstIARF
[highk
layer,fractures)

. Decreasing
slop

p horimntat @
p~= 05

hnsitiO

. More
Decreasing
slop

. Ap * ke-2s
.
St. Line,
or W
. slope =
ti2 (Trans.)or
PD=o.ti, Trans.
o (Pss)
0rpDc0.2.5,pss

;econd IARF
TOM
;ystem)

. similar
slop
toFirst
SARF

p horizontal@
p~=o.5

. St. Line,
. Slope= m
makh
. 4P,~ * s

. St.Line,
. slop =m
m * kh,p.
. zw,h~

;ingk.NOFlow
bmmdary

)utwNo FIOW
:oundaies
lawdown
&ztzOnly)

p horizontal@
p~=1.o

. w.Line,

. Unit Slopefor Ap

.s lope=m*

and p
. Apandp coincide

m*=qAh

* CA
. Pin*

double porosity system bob match the &ta. Geologic 0s seizmic

maps, which may show the presenef of fauka, should be used to
help select tire appropriate model: The B field example illustrates
Uds poiut. Analyses of all field exmnples prezented in thii paper
were made using dugnostic and specialized plots and regression

521

. St. Line,
. slope . h
. Intersectionwith
1~
~.. distanceto
bounday
IncreasingS@e

aualysis. OnIy plotz zhowing the regression auafysis results wiII

be ShOwmto shorten the paper.
B Field Example. The B Field is located in the Lower Wflcox
trend ia .SouthTezas. The reseryoir sanda are part of a deltaic system. hw permeability and thiu sands prevail in this area. Welf X

I
4

SPE 28008

USE OFTRANSIENT TESTING IN RESERVOIR MANAGEMENT

was drilled m z depth of about 13,fW3f&i encouirtering abbut 15

feet of potenti~ly productive pay. A pressure buildup te.vtwas run
prior to completion in order to identify reservoir characteristics.
The well ROWW10f21hours at 385 MCWD and was then shut-in
for a 72 hour pmasnre buildup test.T& shape of the pressure derivative curve indicated that two r&ervoir models could be used to describe the reservoir. These two
models are the homogeneous reservoir with imemecting boundaries snd the double porosity reservoir.
The data were first analyzed using the homogeneous reservoir
with wellbore storsge and akh nrcdel. The type curve match for
tfds snsfysis is .sbown io Figure 1. This anafysis indicated a permetillity of 0.04 md, a skin of-1.2, an ioitkd reservoti pressure
of 10190 pii, &d two iifimecting boundaries lucated about 25
feet from the weff. These bounties
cotdd not be confirmed by
seismic data which may be due to the resolution of seismic information. However, geologic interpretation supports the pomib~hy
of stmtigraphic boundaries.
The dsta were afso snalyzed using the double porosity r~ervoir
witi wellbore storage aud skin model. The type curve match for
thk analysis. is shown in Figure 2. Thk model indicated a permeabltity of 0.01 md, a skin of:3.6, imdan initial reaemo$ pressure
of 10430 psi, with a Lumbda of 0.0W4 aud an Omega of 007.
Afthough the match with the double porosity model is better than
that with the homogeneous model, both models resulted in a rea-:
smrable match to the pressure data. It ia dfiicirlt to detemrirre
which is the comect mcdel without addhional geological data.
The cunciiir with the homogeneous reservoir ~itfr intersecting
bormdaries model is that the two boundaries must be located very
close to the well. The concern with the double porosity model is
that it isnot fetible for this ssndstone rerervoir. However, high
pemmability streaks in the sandstone layers maybe causing the
double porosity behavior. Addhionel geological information is
needed to deterndne the pmpx reservoir model. fMs decision is
critical since quantitative results from the trsnsient test wonfd be
used for stimulation decisions, well swing determination and
evaluating the ptrmeahility trends for fotnie explorationin the
area.
Permeability
WelI testing is me man method for determining the reservoir permeability. The permeability vahrcr ob-tined from wel tests are
the effective permeabilities of the ffoiiig fluids under reservoir
tempmtnre arrd pressure. The calculated value is the average
permeability over the radius of investigation of the ICSLThefcfore, IJese penneabiities ciiiI be used directly topredict the perforrmiice of the welf. In gas re&fioi& ihe red g~ psendo

522

pteaa-ti is used in tie mmlysis.13 If more than one fluid is fLowing in Orereservoir, the effective permeability of each phaae msy
k cufcnfsted by substituting the flow rste and fluid properties of
that ph&ein one of thefOIIOw~g equations
~ _ 162#B,

~ _ 1637qT,

Klquid) . . . . . . . . . . . . . ...(1)

(&)..................(z)

(liquid) . . . . . . . . . . . . . . . .

k=

4($).,

k=

5.03x104qTp,c pD
~T .
[)Xj m (Gus). . . . . . . . . . . . . . . . . .
SC

Eqi@ions 1 and 2 SK?nsed for semi-log stmigbt line &alysis snd

equations 3 snd 4 arc naed for log-log analysis. When more than
one phase is flowing, the ratio of the effective permeabilities may
bs used to estimate the fluid saturations ssound the weff snd rhe
absolute permcabtity. Relative pemreabfity curves-from core
analysis are needed in this case. Assmue that both oil and water
are being prcduced. From a tr~ent test, we could calcnlate both
k. and ~ and therefore the ratio of k. I ~ = kroi km A plot of
km/ kw versus SWcould then be used tn cdcnlate Sw Knowiug k.
u!& & md the relative permeability characteristic of the rcrer.
VOU,one can cafcufate the absolute permeability vahre k. The
absolute permeability vafue is what should be used in mawvoir
simrdation studies together with the same relative permeability
funCtiOnused in the Smdysis of the welf teSLlMs Way,Consistent
dsta wifl be used and the simulator wilf easily march the well
pressure aud rates. There are special cases of multiphaae flow in
the reservoir where a function of the pressure rather than the
pressure is used in the armfysis. The two most widely used functions sre the gas condensate pseudo pressure frmction14 and the
COSI&gasification pseudo pressure ftinction.15 In both cases, the
saturations in the reservoirs are reduced to be functions of the
pressure. Both the effective rmd absolute permcsbiiities of the
formation may be calcular&l from a single test in these cases.
In homogeneous isotropic reservoirs, there is only one value of
m~eablfity at afl locations and in alf directions. And if such a
reservoir exista, thii single value of permeability wifl be eafcuk$cd tlom single wdl tests. In homogeneous anirmropic reservoirs, the permeabllhy changes with direction. Single welf tests
yield the geometdc memr of the msxim~ snd min@mr penneabiity vafuea,~=n;

SPE 28oo8

M.M. KAi4AL, 6. G. FREYDER AND M. A. MURRAY

Multiple well transient tests (i.e., interference and pulse tests)

may be used in homogenorm anisotropic reservoirs to identify the
permeability aa a function of dimction.!$)17
In layered rderviiiis, the vaiie of kh ealculmed from a single
wetf testis the sum of kh of alf layers, Zkh. Thk is tme if kh is
calculated from the total system bifinite acting radM flow
regime. A goud way to determine tbe vafues of permeabifities of
individual layers fromconventional well tests is to integrate the
results with hose frorir core anafyiis. The pmmeabflity vafues
from core analysis are usually absolute ~r pemreatdliries at
aoonspheric condhions for asmatl sample (inches) of the reservoir. These smples may have been distmted during dritliig or
laboratory prepamtions. Whereas, as mentiooed before, the permeabilities frnm well tests are effective in-situ values averaged
o~er the rad@s of investigation of the test. Therefore, the valu~
from these two methnds witl not be tie same. Core m-afysis prOVides vatues for the permesbitity in each fayer. We should divide
the welt test kb vefue among the various fayccs with the same
proportions indicated from the core analysis. ~k is equivalent
to fixing the perm+~lity ~tfosmnon,gthe ~ffirent layers using
cwe amdysi.v&ta. The assumption
@ i: .tbat we :g~~ivg
.:, ,.. use~
.
penueability fmrctihu5 ~oi efl layers are similar.
The followirig field example illustrates the use of log and geologic data together with transient teats to chmacimize a layered
reswvoir.
Kupsrruk 1E.13 Field Example. The Kuparuk field is Iocmed on
the Alaskan North Slope. The reservoir consists of two distinct
sands which are Lowsr Cretaceous, shallow, marine shelf sand
dcposita separated by a major uncunfo~ty. ~$ upper sand is
designated the C sahd and the lower the A sand. The reservoir is
atso highly fanlted with throws up to several hundred feet.ls Well
IE.13 was dritled in late 1982 and plac~ on water inj.@on in
early 1983 with inj~tion in the A aamf only. In mid 1988 the C
sand was performed aud comingled injeclion in both the A and C
W@ began. In order to properly interpret two fayef systems, as
much information as possible should k obtained from other reservoir description sources. Therefore, log interptation of the
two zones waa used and it showed the C sand to have a net pay
tfickness of 34 feet with a pwosity of 21% and the A sarrd to
have a net pay thiikneas of 26 feet witi a porosity of 15.8%.
Geologic infommtion indicates that faulting exists .on atl sides
ranging from 190 feet to 750 f&itfrom the &elk tiotiiver, it is not
known whether of not the faults are seahg. In order to effectively manage the watertfood, a pressure fafloff test waa conducted to detemrinri reservoir prcssme and to determine if the
faults are aeahng. Since this well is injecting into two sepamted
sanda, a two layered reservoir model was used. The initiat guess
for the vatues of permcabtitiea was obtained using the kh fmm
an earlier tmrrsient test when the well was completed in the A
523

aand only, aad the total kh for the A and C sands from this test.
Anafysis of the data as shown in Figure 3 resulted in an effective
water permeabtity of 36.9 md for the C sand end 15.6 rnd-for the
A saod with a sdmnlated wellbore cnndhion waa verifed in each
zone. Sealing fardta were recognized from the pressure fatloff ttit
in a pamflel orientation to tie wellbore at distarrcea of approximately 275 feet and 3Q0feet (i.e., the welf is almost in the middle
of a channel).
fn some cases a speciaJ type of tes~ called Layered Reservoir
TesL19~omaybe USedm dek~fie tie VSJES of h of indivirfuaJ
layers dircctfy ffom tmnsient well testing. The key difference in
these tcsta ia the measurement of the transient flow rates maddition to rhe transient pressnreduring the test. When tie flow mtea
flom individual layers are high enough for meaaarements from
flow meters to be accurate, Uayered Rmmvoir Testing provides
an exceflent methnd to estimate tbe pemreabilities of individual
layers

Skin
Skin isrme of the vkablea ~atchanges with dme and sometimes

the flow rate, is best identified from well testing. The skin vefrre
cafcufated form the infinite acting md~ flow data is the total skin
which may include sevcmf components. The following cqaatian
incIudes some of the componenra that may contribute to the totat
ti]n.
s = Sd+ Sperf $pp+slgrb

+sfmc . . . . . . . . . . . . . . . . . (5)

Of the above mentioned components, we may be able to cafadate

only tbe vafucs of Sfrac (skin on face of fractnsc) and Sruti (the
non-Darcy stin) dwectfy form the well test data. Some of the
olJrer .componentr may be estimated from available corrcfa~ons:l-n
The fractuse face akin can be estimated from the difference
LxXweenthe meaaured pressure and the exuwpotated preaaurc
tiom the linear flow portion inthe sqmue mot of time plot at the
beginning of the test.n
The non-Darcy (rate acnaitive) akin can be calculated by mnning
at hat two teats at two dlffcrent rates, plotting the tntat skin versus rate on a Carfeaian plot mrd extrapolating the resulting
?,Waight
~me tu a flow rate value of zero. The skin vafue at the
.flow.mte valre of zero is the damage skin. The turbulent skin at
any rate is the difference between the total skbr and the damage
skisr.~
Skin vafues caIcufated from well tests abould be used to identify
flow damage during ebifling, completion, production or injection
and determine tie need for workover or stirmdatiorr of the wells.

SPE 28008

USE OF TR+iNSIENT TESTING IN RESERVOIR MANAGEMENT

The best estimate of the skin should be used in predicting tie

well performance and ~ktofy matctilng t!re yell p~duction rates
aud pressnres in simulation studies. If the simulator being used
allows for specifying tie vulres. of sW, fien such af&hue
should be used. Cam must betaken to make sure the skin is not
accounted for ttike or ignored. For ex@ple, if part of tbe skin is
due m partial penetration aud tie well is iiiulated as a partially
penetrating well, then SPP should not be included in the skin
value assigned. to tie well. ,$pPwill be autonraticalfy accmmted
for by the flow regimes iu the simulator. Tbe same. tigument
holds for fractures, multiphase tlow.in gas condensate resr?moim,
etc. If the simnlator. cannot handle the specitied value of the skin,
then ita effect maybe modeled by akecing the permeability in the
well grid block and snrrourcling grid blocks usiug the Hawkins
formrda.zs A@in, care mu~t be taken ii these c~es toavoidwing
unrealisticpermeability values. For exmnpl% in case of a large
negative s~n, the engineer may find that a very high value (even
a negative value) should Lwused for the permeability. Irr we case
of a large positive skin, a very small value of permeability may
be indicated. In these cases the number of grid blocks used for
altering the permeabdity should be iucreased (increasing r, in
Eqiiation 6) iuitif realistic ~~bflity
values am obt@ed.

$=(;-,)+]
~~~~

~~~~
.

............................. ,0

Table Y Summary of Methcds to CsdcrdateReservoir Presmrs in

the Drainage Area

=~~
26 :

29

Method

Required
I fnfomlation

MBH

k, $ II>Cl>A

Muskat

Reservoir
Shape

Time
Range

Many
shapes

s;ty~g

Any
sha~

Late time

Tlie ~iee methods tfmt depend&r the infinite acting dial flow
period require the use oftheproducing time tp However, when ~
is much greater tbau the time required to reach pseudo-steady
sure $....$ ~ould be replaced by fw,.31 Forsimulation stdia,
the reservoir pressnre that is needed from welf tests, is the pressure in the well grid block. Peaceman presented methods to @:
culate tie grid block pressures.32-M Pcacemau abowed that the
well block pressure., Pa is related to the flowing pressure. in the
well Ptiamd the reservoir propaties by the following equations

Reservoir Pressure
Reiervoir preasrire is one of the primary parametes needed to
pro@y manage the reservoir. It is one of the pammetara that
change whh rime. Monitoring the pressom is essential whether
the reservoir m~ageinent method is a simple material balance
calculation or a complicated reservoir simulator. When a well is
drilled, the hritiaf reservoir pressuce can beat be obtained from tie
first shut-in periti of a drillstem test or from the preSSuiewOfiIe
of a repeaUmuMple formation test. If the well is a discovery well,
tbeu tJreinitiaf pressure is afso a key parameter in identifying the
original hydrocarbons in place.

-ln~.

0=f+%w

.,...,.....,..........,;,.

(-/)

where
r. = 0.14 (A.r2 + Ay2) 1/2 . . . . . . . . . . . . . . . . . ...-........(8)
for isotropic reservoirs,:
and

For materird bdauce calculations, one average reservoir pmssurc

should be computed for the whole reservoir at every tirue step.
The avefage pressu~-?,hould be crdcukrtezlby avemging the reservoir pressures witMm the drainage areas of itrdivi.dua! wells.
Severti metfrudi are used to obtain ihw pressures. MI average
reservoir pressure methods are for closed systems and they use
simple material bafance concepts with a given r~ervoir description and flow rate history. Table 3 summarizes these methods, the
data required for each method, the reservoir shapes where the
method is applicable, aud the time rauge of ihe neidid &ta..

0.284[ (k /Q 1/2Ax2 + (!X /k ) 1/2Ay2] 112

i-. = ~
, ,.=(9)
(ky /:x) 1(4+ (kx / k)) 1{4
for arrisotropic resmwoirs.

524

M. M. KAMAL, D. G. FREYDER AND M. A. MURRAY

SPE 28008

Fracture Length and Conductivity

The Iengtfr and conductivity of a hydmnfic fracture that intersect
the weflbore can be estimated from ua.nsient tests. These values
am the effective (notnecessirify the physicaI) length and conductivity of the fractures. The effectiieva@es should be.psed so p%
diet welf and reservoir performance rmd to manage field
operations. The r&V,bnis Ihatfor a given reservoir and fiactnre
conductivity incsessing the physieaf flacture length beyond a eertairr vrdue dmmiot improve rhe productivity of the well. The
effeetive welf raclus for a @@r@ well ~.a bfmgded reservoirs
independent of the fracture length beyond a certain vahre.35This
effective well sadius ,is given by
~;=o.357*

kw

.. . . .. . . . .. . ... . . ... . . . . . ..... ..... . . . . . . . . ....(lo)

For steady stare fluid flow, Ure maximum effective half fracture
length is
.
.. . .. ........ . . ... ,,
...
..

Jm.. =

1.91q

. .... T.-. . . ... ... . ..=.... T.. . . .. . . . . .

(11)

Appendw A presents the source of tire above two equations.

ReseNoir

Heterogeneity

Use of mansient tests to help characterize dKfererit aspects of reservoir heterogeneity has been the subject of several publiea~ions~.16,3&37
One of the important f;atures of ieservoir description is the presence of outer boundaries that determine the size of the reservoir,
the field reserves and, in some eases, the proper depletion
scheme. The following two field examples show the use of transient tests mgerher with geologic information to identify reser..
voir boundaries.
r.
Amaeker-Tippett
Kincaid 3 #1 Field Example. Kin@id 3 #1 is
lccated approximately 50 miles south of Mldkurd, Texas in tbe
Amacker-T1ppett Field. TMs is a new welf drilled inio the Bend
Iiiestone which was deposited in the Midkmd Basin during the
eiwly Pennsylvanian, a time of majo< regional uplift and faulting
in the Permian Basin. Geologic inforrnatiori indicates major sealing faults Ioeated approximately 100 feet to the South and 1700
feet to the North. Minor faulting issnspected to exist perpendicular to the major fault to. the SOUUIbut @unot_be contirmed f~m
seismic data. This well was originally believed to be located in a
fault block isolated from otier producing wells in the area and
was expected to bean oil producer at originaf reservoir presswe
of 4300 psia. Net pay tldckness was detesmiried to be 14 feet wiffr

525

a porosity of 670 after driffing the well. When prrsiuction tests

were conducted, the welf was foand to produce 1.5 MMCF/D of
rekograde gas and 40 STJND of coirdensare at afomrarion pressum of 2770 psia. It was tbwefme concluded that the welf was
producing from a gas cap. The reservoir pressnre encountered
was considerably beiow the expected initial reservoir presanm of
4300 pski, indicating deplelion from offset wells. To evaluate the
extent of the gas eap end Ioeations of sealing fanfts, a pressnse
buildup test was conducted. The analysis of the pressure btrldup
test using a homogeneous reservoir model with wellbore storage
anp sf@ras shown in Figure 4 co@mcd a scaling fault located
81fektfmm the well. The &Jysis also contlrrrred tbe existence
of parallel ieiding faults located approximately 204 feet from the
weU. Since the geologic data conilrmed the seafing fault SourfIof
the well, the two paraffel faults were identified to be East aad
West of Ure weff. A negative skin was cafcufated indicating that
the near wellbore was free of damage and maybe slightfy sdmulate~ ~e effective gas fxmrretilfiry was 5.4 md as expected.
The most signidcant rt%ft was the analysis tie ability to compute the distance to the gs3/Oil conract of 590 feeL Unfortunately,
t6ik weii war completed in air unexpected gas eap with a pressure
less than V@ expected. Subsequent sirmdations indicated that
pmdueingtbe gas cap wouldnotsignificantly move Oregas / oiI
contact. Results obtained from tbe pressure buifdup test enabled
the gas reserves to be estimated.
Kupamk 2A-21 Field Example. Generaf and geological inf6inration about the Knparnk field were given earlier irr tbe Knpamk
lE-13 field example section. Wefl 2A-21 was completed afong
the western periphery of the field in the upper or C sand. This
well contains 14 feet of gross C sand tiickness with net pay as fittle as 3 feet. The exact tfdckrress of sand contributing to production is not !mow with certainty. The sand thins to no net pay irr a
well located approximately 2500 feet to the North. Geologic
infomration indicates faults located approximately 680 feef to the
East, 490fegt to the Sou~, and 935 feet to tie WWChowever, it
is not known if tbe faults are setimg. A productivity stndy was
tideriaken to eiafuate the C tid protictivity of thiier sections
.tiong the western periphery of the field. As part of the stndy, a
p~ssure buijdup was conducted sftera fewday flow period with
production avemging 500ST8/D to evafuate wellbore condition
and determine the permestility thickness product. The snsfysis
of tbe pressnre buildup test using a homogeneousreservoir model
yiti wellbore storage and skin as shown in Figure 5 indicated a
inbstarrtiaf wellbore danmge condition and a pemreatility tfdckness product of 9520- rnd-ft..%dysis of the &ta alw indicates
.@wthe weff is Ioea@d ina closed system. B&d u~n rhe available geologic inforiniirion tire faulting to the East, Soirtfr, and
West were determined en be seafing. The fonrtb koundery was
eaferdated to be approximately 2fS30.feet from the well which
coincides wi~ the thinning rif the sand to no net pay to riie North.

-:

~ =

USEOFTRANSIENTTESTING IN RESERVOIR MANAGEMENT

Baaed upon the production test prior to the buildup, the perrneabifity tbickress product, and the high degree of wellbore damage it was concluded that the tbhner C sand intervals are
producible.
In Referenc& 37 the authors state that The data required to select
a proper reservoir management method include interwell reservoir properties, the degree of communication. between dtferent
wells, and inforimition akmt the reservoir heterogeneity. Tram
sierrt-preasure teatirrg hm Men one of the most freqrremfy used
methods for obtaining these data. The results flom transient tests
am integrated with geologic irrfornrarion, cores, logs, and other
data to improve the ovcrafl description of the reservoir.
Of drerrianypieisiire%airsient testing methods avaibble to the
reservoir engineer, multiple-well interference and pulse tests,
have become incfeaaingly popular. Tfis may be attributed to the
relative a%riplicity of ifreoperations rmd tbe rapidity whh whkfr
infornration is obtained compared with mater tests or standard
production operations. Also; significtit improvements in pressnre measurement and re$ording systems; iomputtized &&
acquisition and transmission systems, afong with the recent
development of new ,maqemati@ models of rese~oir ptiormance, am m%ng it fedlble to @ign, conduct, and interpret
multiple-well tests in more reservoirs. Nevertheless, the uniqueness problem associated with reservoir description, espiciily in
beterog.aeous ,systerrra,.requires rjmt all sources of information
be used. These include pressure-transient tests, tracers, and production perfotrimrcedata.

examples were used to show the appropriate use of these tests.

Nomenclature
.
.
.

~
ki
ks
LJ

.
.
.

LJmax
m

.
.

. .

In thk paper swmra.f of tie &rmnonly encountered pitfalls in

using results from trapsient tests werediscussed, &rd aeverrd field
526

fracture pernreabifi~, md
effective permeability for phase i, md
penneabifity of the area affected by skirt, ft [m]
fiaclure haff length, ft. [m]
rnaxinurrn effective fracture half length, ft. [m]
slope of semi-log straight fine psik [lqxd-]
pressure, psi [k Pa]

radius of ar& affected by skin, ft. [m]

wellbore radha, ft [m]
cff.5Ztivewellbore radius, ft [m]
akin, dnensiordess

.
.
.

Several properties if the wellhesefioii syiteiria can be obr&red

from properly desigried, crnrduc.ted and analyzed transient teats.
These properties can dienk used to help rr@rage field ope&ions
mrd hydrocarbon recove~ processes. Since tmrraient testing ia an
inverse problem, the issue of non-uniqueness nray be encomrtered. llerefore, the results frrnrr transient tests should be integrated with tfros.e from other methods to reduce tie non.
nniqueneas problem and obtain the most likely reservoir description. Care should be exercised when compming ieiidti from different methods, Jrrce the sarrre variable rrarne may refer io
dtifeient vafuea ~ diff~ent techniques. E~ples
v .perrneabflity fromcbies~dtnmaient testa, and fiactiie lengths from r6ck
medmnics and @rrsient W+.,

FVF, RTMSTR[rm ms/stocktmrk m3]

formation rMckness, ft [m]

pressure dtference, psi [k Pa]

correction for pressure at beginning of test,
psi [k Pa]
flow tie, B/f3 [m3/d] or MCF/D

4
AfJcon

zkf/2@v, dimensionless
drainage area, ftz [mz]

pmnreaLdlity,md
.

Remarks

a
A
B
h
k

References 36 and 37 present c~mprehensive review of how to

use well testing to help describe various aspects of reservoir heremgeneity.

Concluding

SPE28008

Saturation, fraction
time, hours
correction for time at beginnirg of test, hours
tinre correction at begiming of test, hours
ternperarure, R
ffacture width, ft [m]
gridblock size irr the x duecrion, ft [m]
-. . .....gidblock side in the y dmction, ft [m]
interporosity flow coefficient
viscosity, cp ma,s]

Subscripts
d
D
tic
_m

.
.

min
,.
0

due to darnage
dimensionless.
due to fracturing
nratch point
direction of rn.@rrrum pmrrenbility
direction ofm~ium
perrneabi!hy
oil
pseudo

.+

SPE 28008

M. M. KAMAL, O. G. FREYDER ANO M. A. MURRAY

. ...,

perf
PP
s
SC

= standard temperature and pressure conditions

Orrb
w=
x

= due to turbulence
water
in.cbe x dmrion
. in they dmction

13. A1-Hrrssijy, R. Ramey, H. J., Jr., arrd Crawford, P. B.: lIre

Ffow of Real Gases through Porous Media, JPT (May
196@,624.636, Trans:,AJME,237.
14. Chopra, A. K. and Cmter, R. D.: Proof of the Two-Phase
Sgady-State Theory for Flow Through Poro~ Media,
SP~(Dec. 19%), 603-608.

Acknowledgments
The atitiors tharrk the management of ARCO Exploration arrd
production T621mology for permission to publish tlds work.
TMirks to J. C. Braden, I. M. Buhidirm aird M. R. R<mait for
reviewing rfre paper and making valuable suggestions. Thartks to
B. S. Grotfr, K. D. Krawietz, J. B. Mcflmireon arrd D. C. Swenson for providing the field data. We SJSOthank Ardta Davis for
preparing the marmscript.

1. Earlougher, R. C. Jr.: Advances in Well TestAm@is, Monograph Series, 5PE, Dallas (1977)5.
2.. ;ee, W. J: Well Testing, Textb.mk Series, SPE, Dallas (1982)
Streltsova, T. D.: Well TestinginHeterogeneous Formations,
Exxon Monographs, John Wfley & Soirs (1988).

4.

Home, R. N.: Modem Well TestAnalysis: A Computer-Aided

Approach, Petroway, PaJo Alto, CA (1990).

15. KmnaJ, M. M., and Six, J. L.: Pressrrrc Trarrsient Testing of

Metfrmre Prodtrcirtg Coalbed, SPE Advarrced Te@rrology
Series, Vol 1, No. 1,1993,195-205.
16. KamaJ, M. M.: The Use of %essure Tmrrsienta to Detibc
Reservoir Heterogeneity, JPT (August 1979) 1060-70,
Trans.: AJME. 267.
H. J. Jr.:Inr@ffence Armlysis for Arrisorsopic For17. tiiy,
mations-A Case History, JPT (October 1975) 1290-&
iiitiS.;H,
259.
18. ,Starley, G. P., Masirm Jr., W. H., Weiss. J. L.. and BolJiiz. J.
D.: FuI1-Field Sirrrufati6n for Deve18pmerit Plamring id
Reservoir Mmragement at .Kupamk River Field, JPT
@lylst
1991) 97<-982:,
~ .

References

3.

12. Arrraku, T., artd Home, R. N.: Diicrimirrarion between Res:

tioir Mo&ls in Well Test .@alysis, paper SPE 26426 presertted. at the 1993 Aun&d Techrdcaf Conference mrd
ExhMtion, Houston, October 3-6,1993. -

due to perforations
= due to partiat penetmrion
. akin

19. Kucuk, F., Karakas, M., and Ayestaren, L; Analysis of

.Nrrmltieously Measti. Prcssare arrd S@fa$e Flow Rare
irrTiansient Wafl Testing, JPT (f%bmary 1985) 323-34.
20. Efdig-Ecrmomides, C. A., mrd Joseph, J.: A New Test for
Determination of Jtrdividual tiryer Propertiesh a Muliifaycred Reservoir, SPEFE (September 1987) 261-83.

5.

Sabet,M. A.: Well I%t Analysis, Gulf Ptrblisbing Co. (199T).

6.

Ragbavmr, R: Well TestAnrdysis, Prcmice Hafl{1993)

7.

Matthews, C. S. mrd-Rii~U,-D~G~-Preiiure Buildup aid

Flow Tests in Wells, Monograph Series, SPE, Daflas (1967J
1. (p.123).

8.

Ehlig-l?torromidcs, C.: Use of the Pressure Derivative for

Dlagnosiig Pkssure-Trarrsient Behavior, JPT (October
1988) 1280-82:

9.

Aflain, O. F. arrd Home, R. N: Use of Artificial Intelfigerice

in WeIf-Test Interpretation, JPT(Mticb1990) 34249.

21. Karal@ M., and ~q, S. M.: Semi-i&Myticsd Productivity Models. for Perforated Completions, SPEPE @eImrary
1991) 73-82.
.
22. Kazmni, H., arrd Seth, M. S.: Effect of Arriiotropy and
SIrarificarion on pressure Transient Analysis of WeIfs with
Restricted Flow Entry, JPT (May 1969) 639+$7, Trims.,
AM< 246.
23. Ramey, H. J., Jr.: practical Use of Modem Well Test Arralysis, paper SPE 5878 presenr@ at the 46th Amrual California
.Regional Meerirrg, Long Beach, CA, April 8-9,1976. Also
see SPE Reprint Series No. 14,46-67.
24. Carter, R. D., Miller, S. C. arrd Riley, H. G.VDetermination
0! Stabrlliiti
Gas Well Performarrcc from Short How
Ttits, JPT(June 1963), Trans., AJME, 228 (651-8).

10: Af-Kaabi, A. V., McVay, D. A., and Lee. W. J.: iJdne &
Eype-fiSystcm tOIden~y a Weli-Test Int@g~tiop ~O~eL_.
JPT (MSy 1990) 654,61.

25. Hawkins, M. F., Jr.: A Note on the Skirr Effect, JMns.,

AIME, 207,356-57.

11. AHain, 0. F.fid Hoirzi, O. P.: A Pc%ticaf Ardfickd Imelfii

gence Applitition irr Well Test Jrrtqnetation, SPE 24287
presented at the 1992 WE European Petroleiun Cotirpirter
Conference, Stivtiger,,lvfay 25-27.

26. Matthews, C. S., Brnns, E; &idHazebroe~ P; A Method

- for Determfrtation of Average Pressure in a Bomrded Reservoir, Trans., AfME, 201, 182-91.

527

10

USE OF TRANSIENT TESTING IN RESERVOIR MANAGEMENT

R.: %%essnre Bfildup in WeflSV ~OC.. T~d

World Pet. Cong., The Hauge (lg5i) Sec. II, 503-23. .

27. Horner, Il.

SPE 28008

and

28.Mifler, C. C., Dy&, A. B1 ad Hutchiison, C. A., Jrj The

Estimation of Permeability and Reservoir Pressnre from Bottom Hole Pressnre Bnifdup characteristics, Trans. MNfE,
189,91-104.

~;D . 5 . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . .
..Lf
Tbatis .. . .

29. Russell, D. G.: Exti%sioris%fF%&ure Buifdup .&r&y;~

Me~ods;JPT@&m&r19ti)
i624-36.

,
r;D

30, Arps, J. J. and Smith~A.E.: ~acticai Lke @Botto.n Hole

Pressure BuIdup CorveS, DrilI and Prod. Prac. API (1949)
155-65.
31. Kazemi H.: Detenirining Average Reservoir Preamire from
Pressure B.uildup Tests, Trans.; AJME, 255,55-62.
32. Peacemsn, D. W.: Interpretation of Well-Block Pressures in
Nmrrerical Rcaervoii ,$imuhioi~ SPEJ (June 1978) 183-94.
33.Peaceman, D.
W;: Inte@etition of Wefi-BIock ~e~we$ in
NumericaJ Reservoir ,%niulation with Non-sqrmfc Grid
Blocks and Atdsowopic Permeability, Trans.; AIME, 275,
111-531-43.
.
1

34. Peaceman, ~. W.: hterpretstion of Welfblock PreNres in

Numerical Reservoir Simulation: Part 3-Off-Cents and
Multiple Wels witfrin a Cellblock, SPERE (lv@y 1990)
:..
227-32.
. , ..... .
35. Rirney,HJl Jr., Private Gnuirtinication,Fall 1993;
36. Kemaf, M. M; interference Wd Pulse Tesdrrg - A Review,
JPT(De&mber 1%33)2257-70.
37. Kan@, M. M. and Hegemerr, P. S.: New Developments in
Multiple-Well Testing, SPEFE (March 1988) 159-68.

(A-2)

[)

FZy 1for

.Tz.06
Lf

4wk

~2Lk
Z.~23

. . . . . . . (A-3)
f

Ramsnging the above eqnations leads to

r;D = 13.38k3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (A-4)
k
for

Lf = 1.91k4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (A-5)
The abwe equations are for incompressible, steady state fluid
flow. A sinriiar result holds for transient flow. For example, Figruc 6.5 page g 1 from tfre report by Barker and Rsmey,3g shows
that
lk
L5 &

w-1

( )

~=y=f
Tw

lk W
,for ~~s
0.02 ..,:...
f

(Am

rearranging yields

38. Pmta, M; Ef[elt of .Verticel Fractures on Reswvoir Behav-.

ioc Incompmasible Fluid Case, SPEJ (Jnne 1961) 105-18.
39.. Barker, B. aird~ey,
H. J. Jr.: Transient Flow tb Finite C&rductivity Vertical Fractures, Stanford, lhrive~ity
Report, Departfnent of Petrole~
Engineering .(Febimry
1977).

r~=0.33k~w
for
Lf22,5

Appendix A

Effective Well Radius for a Fractured Well

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (A-7)

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (A-8)

~.e equations for r; (A4 & A-7) are nearly tbe same. An average_value for the constant is 0357 whi@k nrsdin eqnation10in
Uretext.Tbe limit is dlffezent because readiig from the above referenced graphs is approximate, and maybe LwcauseEqnadon A-5
is for steady state and Equation A-8 is for transient flow. Howevef, tbe cmrclnsion is clear There is a maximum effective fracture length. Larger fractures do not affect well performance
lndCSSkf wlk is klC~i!d.

From Reference 38, Figure 11 rWD becomes finem wi~ a for

a> -3.....
.where

528

SPE 28008

M M. KAMAL, D. G. FREYDER AND M. A. MURRAY

,.. ..

10.03

. . . ..=-

CUm mm
. .

.. . .

. .
,..

i ---:

j,m

TYF+
-..

11

.. .. . . .

.. .. . .. . . . . .

. .. . . . . . .. . ..=. _
.:

.:

A
A

0.40 . . - . ...........=....

:*

. ....

~~
A

0,0!
0.0!

0.10

..
.

Figm

...

mmT!nls /. mu Smrdp
. .

i: Uo[d 6 Match WMI H.mwwn.ou.

..d

Ec,..dar!..

Mod.,

.:- .,
TYW

cm. Match
..-..., -..

..

= Mwd-.
Czkulmei
cd.la@Ll

Pmsw.
Miv

. ...=.

milum

2: ~eld

B Match Whh Double PO,OSW

529

Moa,i

.. . . . . . . . . . .

. . . . . . . .. .

12..

USE OF TRANSIENT TESTING IN RESERVOIR MANAGEMENT

SPE 28008

m
10.0 ...........-.. >...& ..-=-_..=.

7,0
~,

:.

----------A
A

0.1
0.1

1.0

10.0
mm TL@

,..-.

Im.o

ynmw

..

-.
fl!!m

.,

3$ KuPmuk

,- =...

lE.;3

Match with TWO Uyers

~=..,,

..,,
me

and Bondadns

..=.... ...

Mcau

., . . . .

curve hutch
. .

10.0

. . .

.,,

!,...,

0,1,
0.1

1.0

10.0

ma

m.m.o

,0(

..- ,Q.
-

,
SPE 28008

M. M. KAMAL, D. G. FREYDER AND M. A. MURRAY

13

moo.on

103.W

,,y
M

10.0

Km

lGCO.O

10

.
FIwe 5:Kwmuk

2A.21Match with Homogene.xs

531

..6

Wauntiaries

MC4eI

,=